makes no real difference at all as the 560 also comes in 2GB variants
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
2 x GTX560 Ti (SLi) or 1 x GTX580
- Thread starter 1dmf
- Start date
makes no real difference at all as the 560 also comes in 2GB variants
Ok, we get that Metro is a little demanding if you want to play at the maximum settings possible, with tons of AA. Now if you'd like to use a different example, as this one's a little haggard now. That doesn't mean Crysis, as it's not 'a modern game.'
I'm very interested to see what you could get in Metro 2033 benchmark with your 560 2GB x 2
(with the same settings)Ok Harmony, here are my results (took me this long for Steam's servers to be quiet enough to re-download Metro).
This is with two GTX560 2GB cards running at 950MHz on the core, and 2004MHz on the memory.
I think this might just show that VRAM isn't as important at 1920x1200 as you make out. I saw regular usage of over 1150-1250MB VRAM, but these results are worse than the GTX560 1GB SLI you posted previously. I have this on a SB i5 2500K setup, running at 4.6GHz, with the memory at 1600MHz.
This is with two GTX560 2GB cards running at 950MHz on the core, and 2004MHz on the memory.
I think this might just show that VRAM isn't as important at 1920x1200 as you make out. I saw regular usage of over 1150-1250MB VRAM, but these results are worse than the GTX560 1GB SLI you posted previously. I have this on a SB i5 2500K setup, running at 4.6GHz, with the memory at 1600MHz.
@Marvin: I really appreciate your post! Your effort of re-downloading Metro is definitely worth it!
Your results are consistent with the GTX560 Ti 1GB x 2 @ 1GHz I posted above.
What I want people to see is the lag spikes caused by vram limitation. Compare yours against the 1GB version - the 1GB SLI setup has at least twice many lag spikes as yours!
These will be handy for quote when I argue about vram limitation in the future, thank you! It also saves my time and money to buy a 6950 1GB to prove my hypothesis.
I guess the GTX560 Ti 2GB x 2 SLI was probably somewhat impaired by driver "optimization". Since you only see 1.2GB vram usage while I see 1.4GB on my GTX580s, it's very likely that the nVidia driver has optimized vram usage for reference GTX560 Ti 1GB specifically (and it may have treated your cards as 1GB cards without noticing that your cards have 2GB vram available).
Your results are consistent with the GTX560 Ti 1GB x 2 @ 1GHz I posted above.
What I want people to see is the lag spikes caused by vram limitation. Compare yours against the 1GB version - the 1GB SLI setup has at least twice many lag spikes as yours!
These will be handy for quote when I argue about vram limitation in the future, thank you! It also saves my time and money to buy a 6950 1GB to prove my hypothesis.
I guess the GTX560 Ti 2GB x 2 SLI was probably somewhat impaired by driver "optimization". Since you only see 1.2GB vram usage while I see 1.4GB on my GTX580s, it's very likely that the nVidia driver has optimized vram usage for reference GTX560 Ti 1GB specifically (and it may have treated your cards as 1GB cards without noticing that your cards have 2GB vram available).
Last edited: